154 Dr. Barton and Mr. Garrett on Vibration Curves 



features after the lapse of days (see column 3). The nature 

 of this indirect evidence will be more fully appreciated in a 

 detailed examination of the various figures and the circum- 

 stances under which they were obtained. 



Results. 



The chief results are exhibited in the thirty-nine accom- 

 panying figures selected from 54 photographs, each showing 

 the motions of both string and belly. On the margin near 

 each figure a few words of explanation are given affording a 

 clue to its object or significance, fuller details are given below 

 in order. The length of the string on the monochord is 

 100 cm., and positions of bowing, &c. are given in terms of 

 this coordinate. 



Figures 1, 2, & 3 show early photographs in which the 

 string was bowed at 90 cm., 93 cm., and 80 cm. respectively. 

 Its frequency was about 120 per second. It was observed at 

 45*5 cm. The leg of the optical lever was at 50 cm., i. e., at 

 the middle of the belly as regards its length, but near the 

 edge as regards width. It rested upon the bare head of a 

 metal drawing-pin stuck in the belly. The string's motion 

 is shown by a two-step zigzag, which is a Fourier series 

 practically to infinity, and that is the case all through for the 

 string when bowed near one end. The belly's motion as 

 shown by the bright line in the lower half of figure 3 has 

 components of frequencies 1 and 3, their amplitudes being 

 respectively 1 and J. 



Figures 4 & 5 show a pair of photographs taken after 

 a short interval, but under like conditions, to test the regularity 

 of working of the apparatus. The frequency of the string- 

 was about 120 per second. It was bowed at 90 cm. and 

 observed at 44 cm. The leg of the optical lever was at 

 49*8 cm. and rested on a glass plate cemented on the belly. 

 It is seen that the characters of the motions are identical in 

 the two cases. In figure 4 both string and belly have small 

 amplitudes, but in figure 5 both have large amplitudes. The 

 components of the belly's vibration are here of relative 

 frequencies 1, 2, and 3, the respective amplitudes being 

 about 1, -J, and 1. 



Figures 6 & 7 show two photographs taken in quick 

 succession under the same conditions. They afford a striking- 

 proof of the regularity of response of string, belly, and 

 recording apparatus, when the impressed forces and conditions 

 are the same. The belly's motion here consists almost solely 

 of components of frequencies 1 and 3. It is also noteworthy 

 that the phase lag of the belly's vibration relative to that of 



